Development of anthrax vaccines focuses on improving methods for producing the anthrax toxin protective antigen (PA). An avirulent strain of B. anthracis, termed BH450, has been engineered to be sporulation deficient and to lack extracellular proteases. Proteins such as PA secreted from this strain are expected to be more easily purified in a homogeneous form. An earlier candidate vaccine that is modified to be less sensitive to proteases is now in a clinical trial organized by Drs. Robbins and Scheerson (NICHD). These investigators have also developed a candidate conjugate vaccine that induces antibodies to a sugar termed anthrose that is unique to the anthrax spore surface and a few other rare bacteria. Evaluation of this conjugate is underway. A third approach to vaccine development uses a bacteriophage display system to simultaneously display several of the anthrax toxin components (V. Rao, Catholic U.). This unique vaccine platform was shown to induce antibodies to each of the displayed antigens. [unreadable] Another approach to immunological defense against anthrax uses passive administration of monoclonal antibodies. Work in collaboration with Dr. Purcell (LID, NIAID) is extending to other antigens the work previously reported for PA. The cloning of antibody genes from immunized animals consistently produces monoclonal antibodies of extremely high affinity, and it has been possible to identify those with strong neutralizing activity. [unreadable] In the area of small molecule therapeutics, we have continued assessment of modified cyclodextrins, and have made additional refinements to the structures that have increased efficacy in the rat toxin challenge model (S. Bezrukov, NICHD). In addition, a high throughput screen was initiated with the Chemical Genomics Group (C. Austin, NHGRI) using the lethal factor fusion protein to beta-lactamase as a reporter of the anthrax toxin internalization process. A number of candidates were identified that are undergoing further evaluation. Refinements and improvements are underway to increase the activity of the beta-lactamase reporter system so that the screen can be repeated and expanded. [unreadable] In our continuing efforts to develop modified anthrax toxin proteins as anti-tumor agents, we have found that the agents that were shown to work well by intra-tumoral injection also work when administered systemically to mice. Studies in cell culture have helped to show that the efficacy for melanoma of the particular combination that contains native lethal factor depends on expression of the cell surface metalloproteinase and possession of a mutation that activates the same MAP-kinase pathway that the lethal toxin inactivates. Most striking was the finding that combining the PA variant requiring activation by matrix metalloproteases with the native lethal factor produced high potency toward a variety of solid tumors in mice. Examination of the molecular basis of this broad specificity showed it to be due to inhibition of tumor angiogenesis rather than direct action on the tumor cells.